I have no idea if the bias is strong enough to profitably trade on it (or whether it even exists).

But your statement seems stronger, i.e. that such a strategy is somehow fundamentally and inherently impossible, so I think it's on you to explain why that is supposedly the case.

For example, assuming a hypothetical consistent "yes" bias of 10%, would you still say it's impossible? Why? Basically, are you saying it's impossible because of the actual observed "yes" bias being too weak or for some other reason?